Results for the local assessment (PhD Chapter 1)

This series of files compile all analyses done during Chapter 1 for the local assessment (Campaign 2014):

All analyses have been done with PRIMER-e 6 and R 3.6.0.

Caracteristics of each campaign

2014 2016 2017
Sampling date August-September June to August July
Criteria for perturbation Potentially impacted if close to the city or industries, References outside the bay Human-impacted if in a region with a highly populated area, industries and maritime activities, Reference if none of these criteria Human-impacted if in a region with a highly populated area, industries and maritime activities, Reference if none of these criteria
Regions considered BSI BSI, CPC, BDA, MR BSI, MR
Number of sampled stations 40 (20 HI, 20 R) 78 (26 BSI, 19 CPC, 18 BDA, 15 MR) 126 (111 BSI, 15 MR)
Parameters sampled Organic matter yes yes yes
Photosynthetic pigments no yes yes
Sediment grain-size yes yes yes
Heavy-metals yes yes (for a limited number of stations) no (interpolated based on 2014 and 2016 values)
Benthic communities Compartment targeted Macro-infauna Macro-infauna Macro-infauna
Sieved used 500 µm 1 mm 500 µm and 1 mm
Conservation technique Formaldehyle Formaldehyle Formaldehyle
Others N.A. N.A. N.A.

Click on the table of contents in the left margin to assess a specific analysis
Click on a figure to zoom it

To assess Section 1, click here.
To go back to the summary page, click here.

We selected variables and characteristic species (using IndVal index and SIMPER procedure, see Section 2) for the analyses:

Workspace preparation

Here, we use data from subtidal ecosystems (see metadata files for more information)
Only stations that have been sampled both for abiotic parameters and benthic species were included.
The script below includes personnal functions, refined data, parameters for each campaign and global means, sd, se.

1. Permutational Analyses of Variance

Results of univariate PermANOVAs on parameters and multivariate PermANOVA on the whole benthic community are presented in the table below.

Variable Condition Site(Co) Significative groups of similar sites (p > 0.05)
depth All sites in the same group
om S S (P1 P2 P3), (P4 R2), (R1 R2 R3)
gravel S (P1 P2 P3 P4 R3 R4), (R1 R2)
sand S All sites in the same group
silt S (P1 P2 P3 P4 R2 R3), (R1 R2), (R1 R4), (R2 R3 R4)
clay S (P1 P2 P3 P4), (P4 R1 R2 R3 R4), (R1 R2 R3), (R3 R4)
arsenic S (P1 P2), (P3 P4 R2), (P3 P4 R1 R3 R4)
cadmium S All except (P1 R2), (P1 R3), (P2 R2), (P2 R3), (P3 R2), (P3 R3)
chromium S (P1 P2 P3 R1 R4), (P4 R2 R3 R4)
copper S S (P1 P2 P3), (P1 P3 P4), (P4 R1 R2), (R1 R2 R3), (R2 R3 R4)
iron All except (P1 R3), (P2 R3), (R1 R3)
manganese S (P1 P2), (P3 P4 R1 R4), (R2 R3)
mercury (P1 P2 P3), (P2 P4 R1 R2 R3 R4)
lead S (P1 P2), (P1 P3), (P4 R1 R2 R3 R4)
zinc S (P1 P2 P3 P4), (P4 R1 R2 R4), (P4 R2 R3 R4)
S (500 µm) S (P1 P2 P3), (P4 R1 R3 R4), (P4 R2 R3 R4)
N (500 µm) S (P1 P2 P3), (P4 R2 R3 R4), (R1 R4)
H (500 µm) All except (P2 P3), (P3 P4)
J (500 µm) All except (P1 P4), (P1 R1), (P2 P3), (P2 P4), (P2 R1), (P2 R2)
Bneo (500 µm) S S (P1 P2), (P1 P3), (P4 R1 R2 R3), (R1 R2 R3 R4)
Ssol (500 µm) S All except (P1 P2 P3 R1 R3 R4)
ALL SPECIES (500 µm) S S (P1 P2), (R1 R4), (R2 R3)

2. IndVal and SIMPER

These analyses allowed to select species as dependant variables for the regressions. We used results from PRIMER to justify further their choice.

##                          cluster indicator_value probability
## bipalponephtys_neotena         1          0.9490       0.001
## prionospio_steenstrupi         1          0.8969       0.001
## nephtys_sp                     1          0.8494       0.001
## phyllodoce_groenlandica        1          0.8337       0.001
## phoronida                      1          0.7986       0.001
## capitella_sp                   1          0.7940       0.001
## scoloplos_armiger              1          0.7828       0.002
## cirratulidae_spp               1          0.7470       0.001
## limecola_balthica              1          0.7465       0.001
## sarsicytheridea_sp             1          0.6974       0.001
## eteone_sp                      1          0.6386       0.001
## hediste_diversicolor           1          0.5500       0.001
## euchone_analis                 1          0.4500       0.003
## pholoe_longa                   1          0.4015       0.030
## pholoe_sp                      1          0.3792       0.019
## pontoporeia_femorata           1          0.3500       0.008
## podocopida                     1          0.3466       0.015
## diastylis_sculpta              1          0.3435       0.017
## glycera_dibranchiata           1          0.3360       0.009
## axinopsida_orbiculata          1          0.3000       0.026
## praxillella_praetermissa       1          0.3000       0.026
## sabellidae_spp                 1          0.3000       0.023
## tharyx_sp                      1          0.3000       0.017
## spisula_solidissima            2          0.7515       0.001
## polygordius_sp                 2          0.7397       0.001
## echinarachnius_parma           2          0.7000       0.001
## 
## Sum of probabilities                 =  67.855 
## 
## Sum of Indicator Values              =  24.26 
## 
## Sum of Significant Indicator Values  =  15.28 
## 
## Number of Significant Indicators     =  26 
## 
## Significant Indicator Distribution
## 
##  1  2 
## 23  3
SIMPER results (average dissimilarity: 96.41 )
  average sd ratio ava avb cumsum
bipalponephtys_neotena 0.272 0.152 1.79 425 0.45 0.282
nephtys_sp 0.222 0.143 1.55 345 0.25 0.513
prionospio_steenstrupi 0.0581 0.065 0.895 58 0.2 0.573
scoloplos_armiger 0.0439 0.0524 0.838 63.5 1.4 0.618
spisula_solidissima 0.0398 0.0919 0.433 1.25 19.4 0.66
phoronida 0.0345 0.0372 0.926 56.9 0.1 0.695
phoxocephalus_holbolli 0.0249 0.0535 0.465 4.65 16.2 0.721
polygordius_sp 0.024 0.0971 0.247 0.5 36 0.746
phyllodoce_groenlandica 0.0207 0.0195 1.06 25.5 0.5 0.768
harpacticoida 0.0207 0.0446 0.463 10.9 10.2 0.789
capitella_sp 0.0198 0.0217 0.911 26.2 0.2 0.81
mytilus_sp 0.0153 0.0636 0.241 0.3 15.9 0.826
oligochaeta 0.014 0.0528 0.265 1.5 4.45 0.84
echinarachnius_parma 0.0136 0.0382 0.356 0 6.8 0.854
limecola_balthica 0.0109 0.0175 0.621 10.6 0.05 0.865
hediste_diversicolor 0.0105 0.0416 0.252 2.95 0 0.876
pholoe_minuta_tecta 0.0099 0.0374 0.265 4.95 2.75 0.887
glycera_sp 0.00984 0.0295 0.333 1.35 0 0.897

3. Univariate regressions

Independant variables are habitat parameters and heavy metal concentrations, dependant variables are diversity indices and characteristic species abundances.

i) Identification of outliers

To identify stations that are not consistent with the others, we used the multivariate Cook’s Distance (CD) on the uncorrelated variables. A significative threshold of 4 times the mean of CD has been established.
We used linear models for the regressions on diversity indices, and generalized linear models with Poisson distribution for species abundances.

Based on Cook’s Distance, we identified stations 1, 19 and 29 as general outliers. They have been deleted for the following analyses.

ii) Correlations between parameters

Correlations have been calculated with Spearman’s rank coefficient.

Correlation coefficients between habitat parameters and metals concentrations
  depth om gravel sand silt clay arsenic cadmium chromium copper iron manganese mercury lead zinc
depth 1 0.104 -0.054 0.006 -0.017 -0.021 0.025 0.008 -0.168 0.188 -0.127 -0.099 -0.08 -0.078 0.068
om 0.104 1 -0.606 -0.137 -0.439 0.649 0.575 0.288 0.175 0.785 -0.066 0.372 0.702 0.641 0.661
gravel -0.054 -0.606 1 0.236 0.332 -0.754 -0.419 -0.255 -0.162 -0.524 -0.013 -0.384 -0.536 -0.569 -0.607
sand 0.006 -0.137 0.236 1 -0.644 -0.67 -0.327 -0.512 -0.579 -0.415 -0.545 -0.507 -0.297 -0.456 -0.504
silt -0.017 -0.439 0.332 -0.644 1 -0.086 -0.143 0.227 0.345 -0.164 0.418 0.072 -0.233 -0.128 -0.099
clay -0.021 0.649 -0.754 -0.67 -0.086 1 0.602 0.522 0.476 0.707 0.312 0.624 0.67 0.782 0.809
arsenic 0.025 0.575 -0.419 -0.327 -0.143 0.602 1 0.482 0.416 0.681 0.291 0.572 0.584 0.68 0.612
cadmium 0.008 0.288 -0.255 -0.512 0.227 0.522 0.482 1 0.855 0.519 0.725 0.822 0.452 0.792 0.775
chromium -0.168 0.175 -0.162 -0.579 0.345 0.476 0.416 0.855 1 0.448 0.888 0.82 0.445 0.744 0.719
copper 0.188 0.785 -0.524 -0.415 -0.164 0.707 0.681 0.519 0.448 1 0.286 0.587 0.646 0.729 0.837
iron -0.127 -0.066 -0.013 -0.545 0.418 0.312 0.291 0.725 0.888 0.286 1 0.708 0.174 0.58 0.566
manganese -0.099 0.372 -0.384 -0.507 0.072 0.624 0.572 0.822 0.82 0.587 0.708 1 0.591 0.832 0.792
mercury -0.08 0.702 -0.536 -0.297 -0.233 0.67 0.584 0.452 0.445 0.646 0.174 0.591 1 0.728 0.659
lead -0.078 0.641 -0.569 -0.456 -0.128 0.782 0.68 0.792 0.744 0.729 0.58 0.832 0.728 1 0.914
zinc 0.068 0.661 -0.607 -0.504 -0.099 0.809 0.612 0.775 0.719 0.837 0.566 0.792 0.659 0.914 1

According to these results, the following variables are highly correlated so they have been considered together in the regressions (\(|\rho|\) > 0.80):

  • cadmium and chromium concentrations (chromium deleted)
  • lead and zinc concentrations (zinc deleted)

We also decided to exclude clay content in the regressions, as it tends to increase drasticaly VIFs due to a marginal negative correlation with sand.

iii) Simple regressions

These analyses have been done to explore the relationships between variables. As it is a huge number of results to interpret, only multiple regressions will be included in the article.

Adjusted R-squared of simple regressions with all variables (continued below)
  depth om gravel sand silt arsenic cadmium copper iron
S -0.02809 0.3729 0.1572 0.005039 0.2358 0.3219 0.1434 0.4527 -0.02687
N -0.00764 0.4488 0.1816 0.04396 0.2275 0.56 0.183 0.6473 -0.01284
H -0.02798 0.01303 -0.0257 -0.009711 -0.02158 -0.02834 0.01109 0.01352 0.007732
J -0.02213 0.03825 0.02916 -0.02281 0.07349 0.1648 -0.02466 0.06347 -0.01873
Bneo 0.02071 0.3162 0.1303 0.06793 0.1663 0.6162 0.2357 0.6512 0.02372
Ssol -0.02849 0.06797 -0.02856 -0.001817 -0.01211 0.05463 -0.004296 0.1686 -0.01943
  manganese mercury lead
S 0.1356 0.3132 0.4404
N 0.4417 0.2645 0.7273
H -0.02525 0.0008677 -0.02508
J 0.07295 0.01996 0.147
Bneo 0.6945 0.163 0.7548
Ssol 0.01571 0.01922 0.05102
p-values of simple regressions with all variables (continued below)
  depth om gravel sand silt arsenic cadmium copper iron
S 0.8988 3.576e-05 0.008752 0.2843 0.001363 0.0001491 0.01199 3.047e-06 0.8111
N 0.3996 3.466e-06 0.004978 0.1122 0.001671 6.116e-08 0.004812 1.193e-09 0.4658
H 0.8884 0.2326 0.7561 0.4242 0.6277 0.9301 0.2441 0.2299 0.2655
J 0.6415 0.1279 0.158 0.6597 0.05756 0.007336 0.717 0.07208 0.5646
Bneo 0.193 0.0001738 0.01611 0.06521 0.007092 5.356e-09 0.001367 9.836e-10 0.1797
Ssol 0.9586 0.06515 0.9855 0.3403 0.4555 0.08799 0.364 0.006732 0.579
  manganese mercury lead
S 0.0143 0.0001886 4.546e-06
N 4.369e-06 0.0006679 1.263e-11
H 0.7384 0.3168 0.7321
J 0.05827 0.1965 0.01103
Bneo 9.365e-11 0.007656 1.927e-12
Ssol 0.2178 0.2001 0.09549

iv) Multiple regressions

This section presents analyses done (i) to determine which model (metals, parameters or all) decribes the best the parameters and (ii) which variables are the most important to explain the parameters.
We used linear models for the regressions on diversity indices, and generalized linear models with Poisson distribution for species abundances. We also used ZIP models, but they are “computationally” singular, so they have not been computed here.

A. Best model selection

The aim is to descriminate the effect of habitat parameters and heavy metal concentrations on the dependant variables.

Results of the model selection are summurized below (according to AIC).

Model S N H J Bneo Ssol
Full model 223.9 541.4 34.85 -29.91 793.8 415.6
Parameters 236.4 566.5 45.17 -28.33 3222 621.4
Metals 218.5 541.9 32.56 -33.98 4376 516.3

Species richness

  n df AIC ∆AIC R2adj
Metals 37 9 218.5 0 0.67
Full model 37 14 223.9 5.45 0.65
Parameters 37 7 236.4 17.94 0.45

Total abundance

  n df AIC ∆AIC R2adj
Full model 37 14 541.4 0 0.81
Metals 37 9 541.9 0.5559 0.79
Parameters 37 7 566.5 25.12 0.57

Shannon index

  n df AIC ∆AIC R2adj
Metals 37 9 32.56 0 0.25
Full model 37 14 34.85 2.292 0.27
Parameters 37 7 45.17 12.61 -0.09

Piélou’s evenness

  n df AIC ∆AIC R2adj
Metals 37 9 -33.98 0 0.18
Full model 37 14 -29.91 4.072 0.15
Parameters 37 7 -28.33 5.651 0

Abundance of B. neotena

  n df AIC ∆AIC Pseudo-R2
Full model 37 13 793.8 0 0.95
Parameters 37 6 3222 2428 0.8
Metals 37 8 4376 3582 0.73

Abundance of S. sollidissima

  n df AIC ∆AIC Pseudo-R2
Full model 37 13 415.6 0 0.66
Metals 37 8 516.3 100.8 0.57
Parameters 37 6 621.9 206.4 0.48

B. Significative variables selection

We identified which variables are selected after an AIC procedure to best predict the variation of the parameters.

All variables

Results of the selection are summurized below (according to AIC).

Variable S N H J Bneo Ssol
depth - - - - +
om + + -
gravel - -
sand - -
silt - -
arsenic + -
cadmium + + + + -
copper + + + - -
iron - - - - +
manganese - + +
mercury + - -
lead + - - + +
Adjusted-R2 0.68 0.84 0.39 0.28
McFadden Pseudo-R2 0.95 0.66
Species richness
## FULL MODEL
## Adjusted R2 is: 0.65
Fitting linear model: S ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 14.37 5.136 2.797 0.009989 * *
depth -0.4914 0.5206 -0.9439 0.3546
om -0.706 1.16 -0.6087 0.5485
gravel -84.46 98.6 -0.8566 0.4002
sand -0.5724 3.131 -0.1828 0.8565
silt -160.5 201.7 -0.7957 0.434
arsenic 1.627 1.451 1.121 0.2732
cadmium 90.57 34.94 2.592 0.01598 *
copper 0.2321 0.2459 0.9441 0.3545
iron -0.0001452 6.149e-05 -2.362 0.02665 *
manganese -0.002383 0.001405 -1.696 0.1027
mercury 122.8 72.53 1.693 0.1034
lead 0.2523 1.189 0.2121 0.8338
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.16 2.46 1.39 1.64 1.71 2.62 2.19 3.04 2.2 2.11 2.05 4.93 
## REDUCED MODEL
## Adjusted R2 is: 0.68
Fitting linear model: S ~ gravel + arsenic + cadmium + copper + iron + manganese + mercury
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 9.571 2.703 3.541 0.001369 * *
gravel -122.5 73.63 -1.664 0.1069
arsenic 1.935 1.034 1.872 0.07134
cadmium 82.08 30.53 2.688 0.01177 *
copper 0.2431 0.1403 1.732 0.09389
iron -0.0001367 4.825e-05 -2.834 0.008283 * *
manganese -0.001974 0.001306 -1.512 0.1413
mercury 96.68 38.5 2.511 0.01786 *
Variance Inflation Factors
  gravel arsenic cadmium copper iron manganese mercury
VIF 1.08 1.95 2 1.81 1.8 2.05 1.14 
## Analysis of Variance Table
## 
## Model 1: S ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + 
##     iron + manganese + mercury + lead
## Model 2: S ~ gravel + arsenic + cadmium + copper + iron + manganese + 
##     mercury
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     24 431.80                           
## 2     29 478.15 -5    -46.35 0.5152 0.7621
## RMSE for the full model: 5.362224
## RMSE for the reduced model: 5.139873

Total abundance
## FULL MODEL
## Adjusted R2 is: 0.81
Fitting linear model: N ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 445.8 374.9 1.189 0.246
depth -68.25 37.99 -1.796 0.08505
om 158.2 84.65 1.869 0.07381
gravel -4438 7196 -0.6167 0.5432
sand -79.45 228.5 -0.3477 0.7311
silt -3390 14719 -0.2303 0.8198
arsenic 88.48 105.9 0.8358 0.4115
cadmium 861.4 2550 0.3378 0.7384
copper 10.04 17.94 0.5596 0.5809
iron -0.006653 0.004487 -1.483 0.1512
manganese 0.04181 0.1025 0.4077 0.6871
mercury -4355 5294 -0.8227 0.4188
lead 103.6 86.8 1.194 0.2443
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.16 2.46 1.39 1.64 1.71 2.62 2.19 3.04 2.2 2.11 2.05 4.93 
## REDUCED MODEL
## Adjusted R2 is: 0.84
Fitting linear model: N ~ depth + om + iron + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 441.9 266.8 1.657 0.1073
depth -53.72 30.33 -1.771 0.08608
om 97.4 40.34 2.414 0.02165 *
iron -0.00713 0.002258 -3.158 0.003457 * *
lead 203.8 23 8.859 4.027e-10 * * *
Variance Inflation Factors
  depth om iron lead
VIF 1.01 1.28 1.21 1.42 
## Analysis of Variance Table
## 
## Model 1: N ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + 
##     iron + manganese + mercury + lead
## Model 2: N ~ depth + om + iron + lead
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     24 2299972                           
## 2     32 2574319 -8   -274347 0.3578 0.9325
## RMSE for the full model: 390.2629
## RMSE for the reduced model: 318.3117

Shannon index
## FULL MODEL
## Adjusted R2 is: 0.27
Fitting linear model: H ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 2.258 0.3991 5.658 7.938e-06 * * *
depth -0.08271 0.04045 -2.045 0.05202
om -0.05266 0.09012 -0.5843 0.5645
gravel -2.401 7.662 -0.3134 0.7567
sand -0.1723 0.2433 -0.7082 0.4856
silt -6.874 15.67 -0.4387 0.6648
arsenic 0.01703 0.1127 0.1511 0.8812
cadmium 10.45 2.715 3.848 0.0007723 * * *
copper 0.0278 0.0191 1.455 0.1586
iron -1.361e-05 4.778e-06 -2.848 0.008882 * *
manganese -5.912e-05 0.0001092 -0.5415 0.5931
mercury 3.229 5.636 0.573 0.572
lead -0.1287 0.09242 -1.392 0.1766
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.16 2.46 1.39 1.64 1.71 2.62 2.19 3.04 2.2 2.11 2.05 4.93 
## REDUCED MODEL
## Adjusted R2 is: 0.39
Fitting linear model: H ~ depth + cadmium + copper + iron + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 2.07 0.2826 7.324 3.041e-08 * * *
depth -0.07949 0.0354 -2.245 0.03201 *
cadmium 10.38 2.36 4.399 0.0001191 * * *
copper 0.02709 0.01402 1.932 0.0625
iron -1.34e-05 3.418e-06 -3.921 0.000455 * * *
lead -0.1254 0.04629 -2.709 0.01089 *
Variance Inflation Factors
  depth cadmium copper iron lead
VIF 1.11 2.09 2.44 1.72 2.71 
## Analysis of Variance Table
## 
## Model 1: H ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + 
##     iron + manganese + mercury + lead
## Model 2: H ~ depth + cadmium + copper + iron + lead
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     24 2.6071                           
## 2     31 2.7996 -7  -0.19248 0.2531 0.9661
## RMSE for the full model: 0.5229417
## RMSE for the reduced model: 0.4107588

Piélou’s evenness
## FULL MODEL
## Adjusted R2 is: 0.15
Fitting linear model: J ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.9354 0.1663 5.623 8.66e-06 * * *
depth -0.0254 0.01686 -1.507 0.1449
om -0.02181 0.03756 -0.5806 0.5669
gravel 0.6201 3.193 0.1942 0.8476
sand -0.1 0.1014 -0.9864 0.3338
silt -0.6882 6.531 -0.1054 0.917
arsenic -0.03036 0.04698 -0.6462 0.5243
cadmium 2.628 1.131 2.323 0.02898 *
copper 0.005011 0.007962 0.6294 0.5351
iron -3.264e-06 1.991e-06 -1.639 0.1142
manganese 2.094e-05 4.55e-05 0.4602 0.6495
mercury 0.2347 2.349 0.09993 0.9212
lead -0.04441 0.03852 -1.153 0.2603
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.16 2.46 1.39 1.64 1.71 2.62 2.19 3.04 2.2 2.11 2.05 4.93 
## REDUCED MODEL
## Adjusted R2 is: 0.28
Fitting linear model: J ~ depth + cadmium + copper + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.7725 0.1145 6.746 1.276e-07 * * *
depth -0.02427 0.01462 -1.659 0.1068
cadmium 1.803 0.6808 2.649 0.01243 *
copper 0.01051 0.005682 1.85 0.07358
lead -0.06865 0.01939 -3.541 0.001247 * *
Variance Inflation Factors
  depth cadmium copper lead
VIF 1.09 1.43 2.35 2.69 
## Analysis of Variance Table
## 
## Model 1: J ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + 
##     iron + manganese + mercury + lead
## Model 2: J ~ depth + cadmium + copper + lead
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     24 0.45286                           
## 2     32 0.51214 -8 -0.059276 0.3927 0.9137
## RMSE for the full model: 0.1851489
## RMSE for the reduced model: 0.1451642

Abundance of B. neotena
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.95
Fitting generalized (poisson/log) linear model: Bneo ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 7.554 0.1036 72.93 0 * * *
depth -0.2582 0.01164 -22.19 3.893e-109 * * *
om 0.02928 0.01482 1.976 0.04811 *
gravel -69.78 36.35 -1.92 0.05488
sand -6.042 0.6579 -9.183 4.206e-20 * * *
silt -579 36.51 -15.86 1.257e-56 * * *
arsenic 0.01623 0.03112 0.5216 0.602
cadmium 14.89 1.001 14.87 5.351e-50 * * *
copper -0.05223 0.01303 -4.008 6.125e-05 * * *
iron -4.024e-05 3.11e-06 -12.94 2.687e-38 * * *
manganese 0.0003128 2.184e-05 14.32 1.617e-46 * * *
mercury -2.378 0.8345 -2.85 0.004379 * *
lead 0.2258 0.02785 8.107 5.188e-16 * * *
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.3 1.85 1.15 1.13 1.12 3.89 2.76 7.24 4.78 3.14 1.68 6.4 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.95
Fitting generalized (poisson/log) linear model: Bneo ~ depth + om + gravel + sand + silt + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 7.561 0.1024 73.85 0 * * *
depth -0.2571 0.01141 -22.53 2.261e-112 * * *
om 0.02741 0.01437 1.907 0.05655
gravel -70.17 36.7 -1.912 0.05587
sand -6.067 0.6549 -9.264 1.968e-20 * * *
silt -583.1 35.62 -16.37 3.12e-60 * * *
cadmium 14.97 0.9887 15.14 8.52e-52 * * *
copper -0.05682 0.009594 -5.922 3.175e-09 * * *
iron -3.95e-05 2.75e-06 -14.36 8.795e-47 * * *
manganese 0.0003174 2.005e-05 15.83 1.923e-56 * * *
mercury -2.337 0.831 -2.813 0.004914 * *
lead 0.2362 0.01948 12.12 7.814e-34 * * *
Variance Inflation Factors
  depth om gravel sand silt cadmium copper iron manganese mercury lead
VIF 1.27 1.79 1.15 1.13 1.1 2.73 5.34 4.23 2.88 1.67 4.48 
## Analysis of Deviance Table
## 
## Model 1: Bneo ~ depth + om + gravel + sand + silt + arsenic + cadmium + 
##     copper + iron + manganese + mercury + lead
## Model 2: Bneo ~ depth + om + gravel + sand + silt + cadmium + copper + 
##     iron + manganese + mercury + lead
##   Resid. Df Resid. Dev Df Deviance
## 1        24     626.46            
## 2        25     626.73 -1 -0.27254

Abundance of S. solidissima
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.66
Fitting generalized (poisson/log) linear model: Ssol ~ depth + om + gravel + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 3.285 0.6535 5.027 4.976e-07 * * *
depth 0.3929 0.04716 8.33 8.091e-17 * * *
om -0.5295 0.2595 -2.041 0.04128 *
gravel 3.338 8.291 0.4026 0.6873
sand -1.001 0.3568 -2.806 0.005016 * *
silt -110.2 28.85 -3.821 0.0001332 * * *
arsenic -0.4065 0.1486 -2.735 0.006247 * *
cadmium -31.21 4.479 -6.968 3.211e-12 * * *
copper -0.4297 0.05225 -8.223 1.977e-16 * * *
iron 1.842e-05 6.18e-06 2.98 0.002882 * *
manganese 0.001142 0.0004457 2.562 0.01042 *
mercury -1376 85573 -0.01608 0.9872
lead 0.6198 0.1705 3.635 0.0002775 * * *
Variance Inflation Factors
  depth om gravel sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.43 1.72 1.44 3.25 2.61 1.35 4.11 1.69 4.77 3.11 1 3.69 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.66
Fitting generalized (poisson/log) linear model: Ssol ~ depth + om + sand + silt + arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 3.311 0.6496 5.097 3.453e-07 * * *
depth 0.3879 0.04543 8.537 1.376e-17 * * *
om -0.552 0.2537 -2.176 0.02957 *
sand -0.939 0.3206 -2.929 0.003404 * *
silt -106.4 27.11 -3.926 8.626e-05 * * *
arsenic -0.4048 0.1485 -2.727 0.006397 * *
cadmium -31.3 4.497 -6.959 3.422e-12 * * *
copper -0.4209 0.04714 -8.93 4.27e-19 * * *
iron 1.955e-05 5.531e-06 3.534 0.0004088 * * *
manganese 0.001133 0.0004463 2.539 0.01112 *
mercury -1369 85977 -0.01592 0.9873
lead 0.5843 0.1462 3.997 6.411e-05 * * *
Variance Inflation Factors
  depth om sand silt arsenic cadmium copper iron manganese mercury lead
VIF 1.38 1.68 2.92 2.46 1.35 4.13 1.53 4.27 3.11 1 3.16 
## Analysis of Deviance Table
## 
## Model 1: Ssol ~ depth + om + gravel + sand + silt + arsenic + cadmium + 
##     copper + iron + manganese + mercury + lead
## Model 2: Ssol ~ depth + om + sand + silt + arsenic + cadmium + copper + 
##     iron + manganese + mercury + lead
##   Resid. Df Resid. Dev Df Deviance
## 1        24     313.05            
## 2        25     313.21 -1 -0.16134

Parameters

Results of the variables selection are summurized below (according to AIC).

Variable S N H J Bneo Ssol
depth - +
om + + + -
gravel - -
sand - - - +
silt - - + - -
Adjusted-R2 0.47 0.57 0 0.07
McFadden Pseudo-R2 0.8 0.48
Species richness
## FULL MODEL
## Adjusted R2 is: 0.45
Fitting linear model: S ~ depth + om + gravel + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 15.71 4.505 3.486 0.001487 * *
depth 0.156 0.5658 0.2757 0.7846
om 1.904 0.7018 2.713 0.01077 *
gravel -54.46 110.3 -0.4939 0.6248
sand -4.34 3.056 -1.42 0.1656
silt -469.4 210.3 -2.232 0.03299 *
Variance Inflation Factors
  depth om gravel sand silt
VIF 1 1.18 1.23 1.27 1.42 
## REDUCED MODEL
## Adjusted R2 is: 0.47
Fitting linear model: S ~ om + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 16.83 2.002 8.408 1.029e-09 * * *
om 1.911 0.6833 2.797 0.008531 * *
sand -4.963 2.713 -1.829 0.07639
silt -519.7 179.1 -2.902 0.006557 * *
Variance Inflation Factors
  om sand silt
VIF 1.18 1.16 1.24 
## Analysis of Variance Table
## 
## Model 1: S ~ depth + om + gravel + sand + silt
## Model 2: S ~ om + sand + silt
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     31 883.42                           
## 2     33 892.29 -2    -8.867 0.1556 0.8566
## RMSE for the full model: 6.049168
## RMSE for the reduced model: 5.752928

Total abundance
## FULL MODEL
## Adjusted R2 is: 0.57
Fitting linear model: N ~ depth + om + gravel + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 1108 390 2.84 0.007892 * *
depth -58.09 48.98 -1.186 0.2446
om 204.8 60.75 3.37 0.002025 * *
gravel -4234 9545 -0.4436 0.6604
sand -587.6 264.6 -2.221 0.0338 *
silt -47307 18209 -2.598 0.01422 *
Variance Inflation Factors
  depth om gravel sand silt
VIF 1 1.18 1.23 1.27 1.42 
## REDUCED MODEL
## Adjusted R2 is: 0.57
Fitting linear model: N ~ om + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 698.1 177 3.945 0.0003934 * * *
om 206.3 60.4 3.416 0.001705 * *
sand -635.2 239.8 -2.649 0.0123 *
silt -51316 15829 -3.242 0.002714 * *
Variance Inflation Factors
  om sand silt
VIF 1.18 1.16 1.24 
## Analysis of Variance Table
## 
## Model 1: N ~ depth + om + gravel + sand + silt
## Model 2: N ~ om + sand + silt
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     31 6620797                           
## 2     33 6971227 -2   -350429 0.8204 0.4496
## RMSE for the full model: 578.4279
## RMSE for the reduced model: 589.1765

Shannon index
## FULL MODEL
## Adjusted R2 is: -0.09
Fitting linear model: H ~ depth + om + gravel + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.713 0.34 5.038 1.925e-05 * * *
depth -0.005766 0.0427 -0.135 0.8935
om 0.03858 0.05296 0.7286 0.4717
gravel 3.324 8.321 0.3994 0.6923
sand -0.1744 0.2306 -0.756 0.4554
silt -7.465 15.87 -0.4703 0.6414
Variance Inflation Factors
  depth om gravel sand silt
VIF 1 1.18 1.23 1.27 1.42 
## REDUCED MODEL
## Adjusted R2 is: 0
Fitting linear model: H ~ 1
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.681 0.06334 26.54 3.127e-25 * * *

Quitting from lines 756-779 (Chap1_article_local2.Rmd) Error in Qr$qr[p1, p1, drop = FALSE] : indice hors limites De plus : There were 50 or more warnings (use warnings() to see the first 50)

## Analysis of Variance Table
## 
## Model 1: H ~ depth + om + gravel + sand + silt
## Model 2: H ~ 1
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     31 5.0308                           
## 2     36 5.3435 -5  -0.31274 0.3854 0.8549
## RMSE for the full model: 0.4925092
## RMSE for the reduced model: 0.3973865

Piélou’s evenness
## FULL MODEL
## Adjusted R2 is: 0
Fitting linear model: J ~ depth + om + gravel + sand + silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.6915 0.1259 5.493 5.227e-06 * * *
depth -0.008357 0.01581 -0.5285 0.6009
om -0.01578 0.01961 -0.8046 0.4272
gravel 2.014 3.081 0.6536 0.5182
sand -0.03153 0.08542 -0.3691 0.7145
silt 4.178 5.878 0.7107 0.4826
Variance Inflation Factors
  depth om gravel sand silt
VIF 1 1.18 1.23 1.27 1.42 
## REDUCED MODEL
## Adjusted R2 is: 0.07
Fitting linear model: J ~ silt
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.5996 0.03013 19.9 1.159e-20 * * *
silt 7.824 3.985 1.964 0.05756
Variance Inflation Factors
  silt
VIF 1 
## Analysis of Variance Table
## 
## Model 1: J ~ depth + om + gravel + sand + silt
## Model 2: J ~ silt
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     31 0.68997                           
## 2     35 0.72117 -4 -0.031196 0.3504 0.8417
## RMSE for the full model: 0.1715101
## RMSE for the reduced model: 0.1450135

Abundance of B. neotena
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.8
Fitting generalized (poisson/log) linear model: Bneo ~ depth + om + gravel + sand + silt
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 8.255 0.06243 132.2 0 * * *
depth -0.2706 0.008114 -33.35 6.594e-244 * * *
om 0.06107 0.006309 9.68 3.666e-22 * * *
gravel -451 109.1 -4.133 3.582e-05 * * *
sand -5.683 0.5787 -9.821 9.113e-23 * * *
silt -455.5 18.52 -24.59 1.588e-133 * * *
Variance Inflation Factors
  depth om gravel sand silt
VIF 1.01 1.01 1.23 1 1.23 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.8
Fitting generalized (poisson/log) linear model: Bneo ~ depth + om + gravel + sand + silt
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 8.255 0.06243 132.2 0 * * *
depth -0.2706 0.008114 -33.35 6.594e-244 * * *
om 0.06107 0.006309 9.68 3.666e-22 * * *
gravel -451 109.1 -4.133 3.582e-05 * * *
sand -5.683 0.5787 -9.821 9.113e-23 * * *
silt -455.5 18.52 -24.59 1.588e-133 * * *
Variance Inflation Factors
  depth om gravel sand silt
VIF 1.01 1.01 1.23 1 1.23 
## Analysis of Deviance Table
## 
## Model 1: Bneo ~ depth + om + gravel + sand + silt
## Model 2: Bneo ~ depth + om + gravel + sand + silt
##   Resid. Df Resid. Dev Df Deviance
## 1        31     3068.9            
## 2        31     3068.9  0        0

Abundance of S. solidissima
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.48
Fitting generalized (poisson/log) linear model: Ssol ~ depth + om + gravel + sand + silt
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 3.099 0.2686 11.54 8.599e-31 * * *
depth 0.2071 0.03422 6.052 1.431e-09 * * *
om -2.787 0.2142 -13.01 1.018e-38 * * *
gravel -33.2 6.21 -5.347 8.947e-08 * * *
sand 0.3625 0.1828 1.983 0.04733 *
silt -31.67 16.75 -1.891 0.05869
Variance Inflation Factors
  depth om gravel sand silt
VIF 1.09 1.19 1.09 1.67 1.8 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.48
Fitting generalized (poisson/log) linear model: Ssol ~ depth + om + gravel + sand + silt
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 3.099 0.2686 11.54 8.599e-31 * * *
depth 0.2071 0.03422 6.052 1.431e-09 * * *
om -2.787 0.2142 -13.01 1.018e-38 * * *
gravel -33.2 6.21 -5.347 8.947e-08 * * *
sand 0.3625 0.1828 1.983 0.04733 *
silt -31.67 16.75 -1.891 0.05869
Variance Inflation Factors
  depth om gravel sand silt
VIF 1.09 1.19 1.09 1.67 1.8 
## Analysis of Deviance Table
## 
## Model 1: Ssol ~ depth + om + gravel + sand + silt
## Model 2: Ssol ~ depth + om + gravel + sand + silt
##   Resid. Df Resid. Dev Df Deviance
## 1        31     533.42            
## 2        31     533.42  0        0

Metals

Results of the variables selection are summurized below (according to AIC).

Variable S N H J Bneo Ssol
arsenic + -
cadmium + + + + -
copper + -
iron - - - -
manganese - + +
mercury + + + -
lead + + - - + +
Adjusted-R2 0.68 0.81 0.3 0.23
McFadden Pseudo-R2 0.73 0.57
Species richness
## FULL MODEL
## Adjusted R2 is: 0.67
Fitting linear model: S ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 9.74 2.839 3.43 0.00183 * *
arsenic 1.22 1.23 0.9917 0.3296
cadmium 76 31.78 2.391 0.0235 *
copper 0.06047 0.1836 0.3293 0.7443
iron -0.0001518 4.911e-05 -3.092 0.004367 * *
manganese -0.002237 0.001332 -1.679 0.1039
mercury 91.3 39.81 2.293 0.02926 *
lead 1.199 0.8254 1.452 0.1572
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 2.29 2.06 2.34 1.81 2.07 1.16 3.53 
## REDUCED MODEL
## Adjusted R2 is: 0.68
Fitting linear model: S ~ cadmium + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 11.76 1.986 5.922 1.53e-06 * * *
cadmium 72.17 30.44 2.371 0.02415 *
iron -0.0001641 4.471e-05 -3.671 0.0009031 * * *
manganese -0.001655 0.00118 -1.402 0.1707
mercury 85.43 38.72 2.206 0.03491 *
lead 1.744 0.4964 3.514 0.001382 * *
Variance Inflation Factors
  cadmium iron manganese mercury lead
VIF 2 1.68 1.86 1.15 2.16 
## Analysis of Variance Table
## 
## Model 1: S ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: S ~ cadmium + iron + manganese + mercury + lead
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     29 488.31                           
## 2     31 505.63 -2   -17.325 0.5145 0.6032
## RMSE for the full model: 5.285439
## RMSE for the reduced model: 4.810919

Total abundance
## FULL MODEL
## Adjusted R2 is: 0.79
Fitting linear model: N ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 165.2 224.7 0.7349 0.4683
arsenic -14.57 97.35 -0.1496 0.8821
cadmium -317.8 2516 -0.1263 0.9003
copper 2.471 14.54 0.17 0.8662
iron -0.008016 0.003887 -2.062 0.04824 *
manganese 0.05214 0.1054 0.4945 0.6247
mercury 4522 3151 1.435 0.162
lead 208.8 65.33 3.196 0.003355 * *
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 2.29 2.06 2.34 1.81 2.07 1.16 3.53 
## REDUCED MODEL
## Adjusted R2 is: 0.81
Fitting linear model: N ~ iron + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 131 135.5 0.9664 0.3409
iron -0.007999 0.002292 -3.491 0.001391 * *
mercury 4475 2920 1.532 0.135
lead 222.2 22.14 10.03 1.484e-11 * * *
Variance Inflation Factors
  iron mercury lead
VIF 1.14 1.14 1.27 
## Analysis of Variance Table
## 
## Model 1: N ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: N ~ iron + mercury + lead
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     29 3059313                           
## 2     33 3091782 -4    -32469 0.0769 0.9887
## RMSE for the full model: 360.5498
## RMSE for the reduced model: 319.5343

Shannon index
## FULL MODEL
## Adjusted R2 is: 0.25
Fitting linear model: H ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.531 0.2302 6.649 2.732e-07 * * *
arsenic -0.004579 0.09974 -0.04591 0.9637
cadmium 8.755 2.577 3.397 0.001997 * *
copper 0.01746 0.01489 1.172 0.2506
iron -1.122e-05 3.982e-06 -2.818 0.008603 * *
manganese -3.502e-05 0.000108 -0.3242 0.7481
mercury 1.49 3.228 0.4615 0.6479
lead -0.08072 0.06693 -1.206 0.2376
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 2.29 2.06 2.34 1.81 2.07 1.16 3.53 
## REDUCED MODEL
## Adjusted R2 is: 0.3
Fitting linear model: H ~ cadmium + iron + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.56 0.1575 9.91 2.028e-11 * * *
cadmium 9.485 2.399 3.953 0.0003844 * * *
iron -1.347e-05 3.357e-06 -4.013 0.0003245 * * *
lead -0.0414 0.02606 -1.589 0.1217
Variance Inflation Factors
  cadmium iron lead
VIF 1.98 1.58 1.42 
## Analysis of Variance Table
## 
## Model 1: H ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: H ~ cadmium + iron + lead
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1     29 3.2109                           
## 2     33 3.4095 -4  -0.19859 0.4484 0.7727
## RMSE for the full model: 0.5468835
## RMSE for the reduced model: 0.3763263

Piélou’s evenness
## FULL MODEL
## Adjusted R2 is: 0.18
Fitting linear model: J ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.6664 0.09368 7.113 7.929e-08 * * *
arsenic -0.02478 0.04058 -0.6107 0.5462
cadmium 2.195 1.049 2.093 0.04523 *
copper 0.004632 0.006059 0.7645 0.4507
iron -1.721e-06 1.62e-06 -1.062 0.2968
manganese 2.67e-05 4.395e-05 0.6075 0.5483
mercury -0.476 1.313 -0.3625 0.7196
lead -0.04718 0.02723 -1.733 0.0938
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 2.29 2.06 2.34 1.81 2.07 1.16 3.53 
## REDUCED MODEL
## Adjusted R2 is: 0.23
Fitting linear model: J ~ cadmium + lead
  Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.6192 0.06332 9.779 2.062e-11 * * *
cadmium 1.482 0.6875 2.156 0.03826 *
lead -0.03668 0.0104 -3.525 0.001232 * *
Variance Inflation Factors
  cadmium lead
VIF 1.39 1.39 
## Analysis of Variance Table
## 
## Model 1: J ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: J ~ cadmium + lead
##   Res.Df     RSS Df Sum of Sq      F Pr(>F)
## 1     29 0.53155                           
## 2     34 0.58408 -5 -0.052528 0.5731   0.72
## RMSE for the full model: 0.1903827
## RMSE for the reduced model: 0.1401076

Abundance of B. neotena
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.73
Fitting generalized (poisson/log) linear model: Bneo ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 4.568 0.06807 67.1 0 * * *
arsenic 0.1102 0.02357 4.677 2.914e-06 * * *
cadmium 8.229 0.9813 8.386 5.01e-17 * * *
copper 0.0708 0.007082 9.998 1.551e-23 * * *
iron -5.904e-05 2.197e-06 -26.87 4.985e-159 * * *
manganese 0.0002352 1.708e-05 13.77 3.68e-43 * * *
mercury 4.775 0.4928 9.69 3.328e-22 * * *
lead 0.2136 0.02066 10.34 4.616e-25 * * *
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 3.12 3.43 5.28 4.08 2.77 1.08 5.99 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.73
Fitting generalized (poisson/log) linear model: Bneo ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 4.568 0.06807 67.1 0 * * *
arsenic 0.1102 0.02357 4.677 2.914e-06 * * *
cadmium 8.229 0.9813 8.386 5.01e-17 * * *
copper 0.0708 0.007082 9.998 1.551e-23 * * *
iron -5.904e-05 2.197e-06 -26.87 4.985e-159 * * *
manganese 0.0002352 1.708e-05 13.77 3.68e-43 * * *
mercury 4.775 0.4928 9.69 3.328e-22 * * *
lead 0.2136 0.02066 10.34 4.616e-25 * * *
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 3.12 3.43 5.28 4.08 2.77 1.08 5.99 
## Analysis of Deviance Table
## 
## Model 1: Bneo ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: Bneo ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
##   Resid. Df Resid. Dev Df Deviance
## 1        29     4218.5            
## 2        29     4218.5  0        0

Abundance of S. solidissima
## FULL MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.57
Fitting generalized (poisson/log) linear model: Ssol ~ arsenic + cadmium + copper + iron + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 4.156 0.2698 15.41 1.51e-53 * * *
arsenic -0.2082 0.1281 -1.626 0.104
cadmium -21.78 3.852 -5.654 1.568e-08 * * *
copper -0.3867 0.03939 -9.816 9.564e-23 * * *
iron 2.099e-06 3.776e-06 0.556 0.5782
manganese 0.001085 0.0003345 3.245 0.001176 * *
mercury -1480 56871 -0.02603 0.9792
lead 0.7039 0.1109 6.344 2.236e-10 * * *
Variance Inflation Factors
  arsenic cadmium copper iron manganese mercury lead
VIF 1.23 3.4 1.32 2.86 2.41 1 2.48 
## REDUCED MODEL (Poisson)
## McFadden's Pseudo-R2 is: 0.57
Fitting generalized (poisson/log) linear model: Ssol ~ arsenic + cadmium + copper + manganese + mercury + lead
  Estimate Std. Error z value Pr(>|z|)
(Intercept) 4.128 0.2628 15.71 1.391e-55 * * *
arsenic -0.2046 0.1276 -1.603 0.1089
cadmium -21.04 3.615 -5.82 5.887e-09 * * *
copper -0.3947 0.0368 -10.72 7.825e-27 * * *
manganese 0.001128 0.0003259 3.459 0.0005416 * * *
mercury -1567 87268 -0.01796 0.9857
lead 0.7299 0.0994 7.343 2.087e-13 * * *
Variance Inflation Factors
  arsenic cadmium copper manganese mercury lead
VIF 1.23 3.18 1.25 2.35 1 2.22 
## Analysis of Deviance Table
## 
## Model 1: Ssol ~ arsenic + cadmium + copper + iron + manganese + mercury + 
##     lead
## Model 2: Ssol ~ arsenic + cadmium + copper + manganese + mercury + lead
##   Resid. Df Resid. Dev Df Deviance
## 1        29     423.81            
## 2        30     424.11 -1 -0.30874

4. Multivariate regression

Independant variables are habitat parameters and heavy metal concentrations, dependant variables are diversity indices and characteristic species abundances.
Sand variables (Csand, Msand, Fsand) and mud variables (silt, clay) were merged to reduced the problem of model overfitting.
This analysis has been done on PRIMER, with a DistLM to identify the variables that explain the most the community variability and with a dbRDA to plot the results.

2014 dbRDA

2014 dbRDA

Elliot Dreujou

2019-07-12